ISEP - DM – Engenharia de Inteligência Artificial
Permanent URI for this collection
Browse
Browsing ISEP - DM – Engenharia de Inteligência Artificial by Author "Azevedo, João Nuno Silva"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- Self-Checkout System for product recognitionPublication . Azevedo, João Nuno Silva; Ramos, Carlos Fernando da SilvaRecently, retail environments have increasingly turned to self-checkout systems as a way to simplify operations, providing a seamless customer experience. These systems reduce the need for manual labor, speeding up checkout times to offer customers more autonomy. There is a growing need for technologies to address the current challenges related to product identification. The integration of deep learning and computer vision into self-checkout systems has the potential to revolutionize product identification. With real-time classification of products, there is no need for manual input from costumers or employees. However, a significant challenge in product identification is the classification of fruits and vegetables, being a hard challenge due to many similarities between them. These technologies offer more efficiency and convenience to customers, enhancing customer satisfaction, reducing employee-related expenses while reducing transaction errors and optimizing the overall efficiency of the retail checkout process. This dissertation explores the development of two neural network models for the task of fruit classification, one of the current challenges related to product identification. The objective of this research is to assess the effectiveness of these architectures in fruit classification. Both networks were trained and evaluated on a fruit dataset under the same conditions. Then the experiments were conducted to compare the classification accuracy and model efficiency of both approaches. This study gives valuable insights into the application of deep learning techniques for image recognition, with potential for broader classification tasks and future work in fine-tuning model architectures for optimized performance.